About This Project

Hyperspectral imaging could reveal biological changes without dyes or labels, offering a new window into tissue health and disease. Yet no open, biologically validated reference dataset of label-free hyperspectral signatures exists for living epithelial tissue. Using living fruit fly (Drosophila) tissue, we will compare healthy, wounded and dying states. Fluorescent validation will link each spectrum to a defined biological state, creating a calibration benchmark for the emerging field.

Ask the Scientists

What is the context of this research?

Hyperspectral imaging records many wavelengths at every pixel, revealing subtle differences in tissue composition and structure without dyes. Previous studies show that label-free hyperspectral imaging can distinguish living cell states and detect programmed cell death. Yet these studies use different specimens and analysis pipelines, and, to our knowledge, no shared open reference links label-free spectral signatures to independently verified states in living epithelial tissue.

The fruit fly Drosophila provides an established model for studying wound repair and patterned apoptosis under controlled conditions. We will use the wing epithelium to build a reference dataset spanning healthy, wounded and dying tissue. Our hypothesis is that wounding and apoptosis generate reproducible changes in tissue structure and composition that produce distinguishable spectral signatures, and that fluorescent markers will independently confirm the biological state represented by each spectrum.

What is the significance of this project?

What does a healthy cell look like through a hyperspectral camera? What about a wounded one, or a dying one? Surprisingly, there is no open, validated dataset to answer these basic questions.

Hyperspectral imaging is advancing rapidly, but the field lacks a critical foundation: trusted, biologically validated reference data. Without it, researchers cannot easily compare results across studies, calibrate new algorithms, or be certain that a spectral difference reflects real biology rather than experimental variation.

This project addresses that bottleneck directly. We will generate the first open reference dataset from healthy, wounded, and apoptotic Drosophila epithelial tissue, validating each state with independent fluorescent markers. The result will be a controlled, reusable benchmark that helps move hyperspectral biology from promising demonstrations toward a more rigorous, comparable, and collaborative field.

What are the goals of the project?

With this project, we will build the first open reference dataset of label-free hyperspectral signatures from defined epithelial tissue states in Drosophila.

We will establish a controlled ex vivo imaging workflow and collect baseline data from healthy epithelial tissue, then generate wounded and apoptotic states under matched conditions and image them in the same way.

After each hyperspectral acquisition, we will use fluorescent markers to independently validate cell boundaries and dying cells so every spectral profile is anchored in a known biological state. We will then process the data, extract and organise the spectral signatures, and assemble an openly shared dataset with clear metadata and documentation.

Our goal is to create a reliable first resource that others can use to compare results, calibrate analyses, and build future tools within one year.